Improving Internet Usability - A Framework For Domain Name Policy Evaluation

[Josh Rowe <josh@xxxxxxxx>]

In a world with 1.5 billion Internet users (and growing), usability and utility 
of the Internet are key.

Domain names are a fundamental part of the Internet's user interface.  End 
users use domain names to assess the credibility of a destination (Nielsen, J., 
1999).  End users spent between 22% to 25% of their time looking at the domain 
name in search engine results (Cutrell, E. & Guan, Z., 2007).

Improving the usability of the Internet depends upon effective domain name 
policy.  The benefits of more usable domain names include: higher sales, higher 
customer satisfaction, higher productivity, and reduced support costs 
(Usability Professionals' Association, 2008).

Domain name policies worldwide vary considerably between, and sometimes even 
within, countries. As a consequence, end users are inconvenienced by 
contradictory domain name policies, diminishing the predictability of an 
entity's domain name, and thus decreasing usability for end users.

A set of criteria has been developed, backed by academic research, with which 
policy makers can evaluate their domain name policies, in order to improve the 
usability of domain names for Internet end users.

There is significant existing research examining either domain names or 
usability in isolation. However, academic research examining the intersection 
of the two is scarce. The research that does exist, in this somewhat new and 
untried field of study, describes domain names as part of the web user 
interface. This foundation concept is built upon by this study. It was 
established that the predictability (and thus usability) of domain names relies 
on effective domain name policy. The importance of effective domain name policy 
is evident in the way that the non-standardised and widely delegated process of 
domain name policy development leads to unpredictable and inconsistent domain 
names. These attributes lead to poor usability, observable in decreased 
productivity, sales, revenues and customer satisfaction, as well as increased 
training and support costs, development time and costs, and maintenance costs.

In order to address the problem of poor domain name usability, a framework for 
domain name policy evaluation is proposed (see below). This new framework seeks 
to address usability and quality concerns by treating the domain name system as 
a user interface. The framework sets out criteria which allow domain name 
policy makers to critically assess domain name policies with end users in mind. 
The framework has the potential to set an international standard for the 
critical evaluation of domain name policy, and become the basis for further 
research. The framework can be used to evaluate the domain name policy for any 
domain name space, regardless of its position in the overall domain name 
hierarchy. Whilst this study focuses on Internet domain names, the framework 
can also be usefully applied to internal or intranet domain names. 

A Framework For Domain Name Policy Evaluation

Criteria for domain name policy evaluation and Example(s)

A. Who are the intended users for the domain name space?  Consider end users of 
the domain name (not the domain name registrants).

e.g. The intended users for the Australian community geographic domain names 
(act.au, nsw.au, nt.au, qld.au, sa.au, tas.au, vic.au, wa.au) are listed as: 
All members of the Australian Community, Business and Industry, Local Community 
Groups/Organisations, Tourism . Local, State/Territory, National and 
International (One City One Site, 2002).

B. How is the domain name space meant to be interpreted by the intended users?

e.g. The '.com.au' domain name space is for Australian commercial entities 
(auDA, 2005).

C. How else could the domain name space be interpreted by the intended users?

e.g. Italy's '.it' ccTLD (IANA, 2008a) can be interpreted as the abbreviation 
for 'Information Technology' or the English word 'it'.

D. Who are the unintended users for the domain name space?  Consider end users 
of the domain name (not the domain name registrants) who are not the intended 
users for the domain name space.

e.g. A Canadian entrepreneur has struck a deal (Richards, 2007) with the 
country of Cameroon (.cm) to redirect mistyped .com domains names to an 
advertising page.  Under the deal, any request for an unregistered .cm site 
will default to an advertising page . as opposed to the company who happens to 
own the .com domain name.  The entrepreneur is said to be negotiating similar 
deals with Colombia (.co) and Oman (.om) which could be confused with '.com', 
as well as Niger (.ne) and Ethiopia (.et) which could be confused with '.net'.

E. How could the domain name space be interpreted by the unintended users?

e.g. The country names of Australia and Austria are sometimes confused (Mail 
Online, 2007).  Similarly, a user not familiar with the domain name system 
could misinterpret '.au' as the ccTLD for Austria instead of Australia.

F. Is the domain name space consistent compared with other domain name spaces 
for the intended or untended users?

e.g. Educational domain name spaces include 'edu.au', 'ac.nz' and 'sch.uk'.  
For the 250 ccTLDs there are 111 'edu', 44 'ac' and 13 'sch' second level 
domain names (Chan, 2008).

G. What other semantic meanings does the domain name space have for the 
intended or untended users?

e.g. VeriSign reportedly (Big Empire, 2007) pays the Pacific Island of Tuvalu 
$2.2 million dollars per annum to operate the .tv domain; targeted at the TV 
and entertainment industry.  Laos has handed over .la to the LA Names 
Corporation; who misinform their web site visitors that '.LA is the official 
internet address for Los Angeles' (LA Names Corporation, 2008, p. 1) when the 
official ccTLD database says otherwise (IANA, 2008b).  Cue Clothing turned to 
the Cocos (Keeling) Islands for a perfect semantic affinity between their 
company name and the .cc domain extension.  Guernsey's .gg ccTLD presents the 
obvious domain extension for horse betting companies.

H. How easy is the domain name space to spell for intended or unintended users?

e.g. Users from the United States expect 'center' where as British users will 
expect 'centre' (Top Hoster Center, 2007).

I. How easy is the domain name space to type for intended or unintended users?

e.g. Consider the input device (for example United States layout keyboard, 
mobile phone keypad, etc) and the end user's proficiency at using the input 
device.  One entrepreneur recently auctioned 264 'circle shape' .com domains 
using the standard QWERTY keyboard (Nottiger, 2008).

J. How easy is the domain name space to say and pronounce for intended or 
unintended users?

e.g. The domain name space 'naturbruksgymn.se' may be simple to say for a 
Swede, but those from other countries may find it difficult to pronounce.

K. How memorable is the domain name space for intended or unintended users?

e.g. A memorable domain name does not necessarily have to be short; consider 
which of the two domain names would more likely be recalled, 
BobJonesSurfShop.com or BJSShop.com (Implied by Design, 2006).

L. How meaningful is the domain name space in the languages and scripts of the 
intended or untended users?

e.g. '.nu' is the country code top-level domain for the Pacific Island of Niue 
(IANA, 2007).  'Nu' means 'now' in Swedish (Google Translate, 2008b) and 'nude' 
or 'naked' in French and Portuguese (Google Translate, 2008a).   

Bibliography

Rowe, J. (2008). Improving Internet Usability - A Framework For Domain Name 
Policy Evaluation, Retrieved from http://domainusability.com/

Contact

Josh Rowe

josh@xxxxxxxx

DomainUsability.com